Oliver via Moodle

[RabbieBurns]

Is there any way anyone knows about that we can give our students access to the Oliver Library pages via Moodle ?

[CyberNerd]

I think that Solftlink are / were working on this. Give them a call.

[RabbieBurns]

Thanks.

 

Ive found this http://moodle.org/mod/forum/discuss.php?d=34570 but its from 2007 so not sure how relevent it will be..

[RabbieBurns]

Spoke to softlink (AU) they said they have nothing for moodle, and have no plans to develop any such thing..

 

Any idea if the one I linked to would work?

[localzuk]

Spoke to softlink (AU) they said they have nothing for moodle, and have no plans to develop any such thing..

 

Any idea if the one I linked to would work?

 

One of their UK engineers stated to me that they were working on Moodle parts about 6 months ago, for the next release of Oliver.

[RabbieBurns]

Any idea when the new release is due, or if there's anything written confirming this?

[CyberNerd]

One of their UK engineers stated to me that they were working on Moodle parts about 6 months ago, for the next release of Oliver.

 

Thats pretty much what they said to me, and it was those rumors that lead us to upgrade to the latest version - only to find that it wasn't there and the AD "integration" requires CSV imports and is actually more of a pita than I could have possibly imagined.

[feizhou]

Thats pretty much what they said to me, and it was those rumors that lead us to upgrade to the latest version - only to find that it wasn't there and the AD "integration" requires CSV imports and is actually more of a pita than I could have possibly imagined.

 

Hi, do you know what the fields are for the cvs imports? I have Oliver forced on me so I have deal with this 1GB of RAM chewing excuse of a database frontend. Looks like I can forget at integrating it into our moodle too...

[localzuk]

Thats pretty much what they said to me, and it was those rumors that lead us to upgrade to the latest version - only to find that it wasn't there and the AD "integration" requires CSV imports and is actually more of a pita than I could have possibly imagined.

 

As far as I'm aware, the version they were talking about hasn't actually been released yet. ie. this integration is in the *next* version, not the current version. The current version is a transition phase moving from Concordance et al to JBOS and SQL.

[feizhou]

As far as I'm aware, the version they were talking about hasn't actually been released yet. ie. this integration is in the *next* version, not the current version. The current version is a transition phase moving from Concordance et al to JBOS and SQL.

 

I don't care. Don't advertise integration with stuff like Moodle when the current release 4.2 does not have it yet. Oliver is on my list of software to avoid if at all possible. The thing runs in spurts. One minute it is instantaneous in response, the next it takes a few minutes to show anything. The entire Jboss setup chews up to 1G of RAM when it is running. I have never seen such a huge program that is supposed to serve a fair amount of users (hopefully I never get anymore than a few dozen connecting to it) such as a library. No wonder the specifications said 2G minimum. The thing is a COMPLETE joke. And unintuitive. Finally got a copy of the manual and so I have import the kids into the system. Now I just have to remap all the old Alice accounts to the new ones and the thing will hopefully never need my attention for the rest of the year. I am so glad there is such a thing as pymssql. I cannot imagine having to 'fix' the loan history in Windows dot-net or not.

[localzuk]

I don't care. Don't advertise integration with stuff like Moodle when the current release 4.2 does not have it yet. Oliver is on my list of software to avoid if at all possible. The thing runs in spurts. One minute it is instantaneous in response, the next it takes a few minutes to show anything. The entire Jboss setup chews up to 1G of RAM when it is running. I have never seen such a huge program that is supposed to serve a fair amount of users (hopefully I never get anymore than a few dozen connecting to it) such as a library. No wonder the specifications said 2G minimum. The thing is a COMPLETE joke. And unintuitive. Finally got a copy of the manual and so I have import the kids into the system. Now I just have to remap all the old Alice accounts to the new ones and the thing will hopefully never need my attention for the rest of the year. I am so glad there is such a thing as pymssql. I cannot imagine having to 'fix' the loan history in Windows dot-net or not.

 

AFAIK, they *aren't* advertising these functions yet! They were presented to people at a 'pre-release' or sneak preview demo. Much like microsoft demoing Windows 7 before it was released.

 

And the reason they use JBOSS and SQL is that it scales very nicely. Whether your school has 3 users using it, or 400, the system can scale to handle it - as that's what those technologies are designed for.

 

Please, don't simply slag off a product because you think it uses too much memory. In this day and age, 1GB of RAM is nothing. We have it running in 2 VMs, the front end has 4GB RAM, and the backend has 3GB RAM. Not had a single issue with it since it went in.

 

You'd be complaining, no doubt, if Oliver used a lot less RAM but then worked poorly...

[CyberNerd]

Hi, do you know what the fields are for the cvs imports? I have Oliver forced on me so I have deal with this 1GB of RAM chewing excuse of a database frontend. Looks like I can forget at integrating it into our moodle too...

 

As we don't have usernames in SIMS we had to export (from SIMS) admission numbers, names and UPN's - we then exported from AD usernames and UPN's (we hold UPN in AD). we could then form a spreadsheet and merge the two together so that the names, UPN's and usernames matched. (the admission number in Oliver is called STCode). The csv can be created from any field - softlink will be able to support with this.

 

I agree, the software is an absolute pile of sh*te.

[localzuk]

As we don't have usernames in SIMS we had to export (from SIMS) admission numbers, names and UPN's - we then exported from AD usernames and UPN's (we hold UPN in AD). we could then form a spreadsheet and merge the two together so that the names, UPN's and usernames matched. (the admission number in Oliver is called STCode). The csv can be created from any field - softlink will be able to support with this.

 

I agree, the software is an absolute pile of sh*te.

 

You shouldn't be using UPN - this number can and does change, and the DfE has always said people should not be using it in third party systems for tracking purposes. Admission Number is more appropriate within a single school.

[feizhou]

AFAIK, they *aren't* advertising these functions yet! They were presented to people at a 'pre-release' or sneak preview demo. Much like microsoft demoing Windows 7 before it was released.

 

And the reason they use JBOSS and SQL is that it scales very nicely. Whether your school has 3 users using it, or 400, the system can scale to handle it - as that's what those technologies are designed for.

 

Please, don't simply slag off a product because you think it uses too much memory. In this day and age, 1GB of RAM is nothing. We have it running in 2 VMs, the front end has 4GB RAM, and the backend has 3GB RAM. Not had a single issue with it since it went in.

 

You'd be complaining, no doubt, if Oliver used a lot less RAM but then worked poorly...

 

Why do you think I would know anything about integration with Moodle IF they did not specifically claim that by moving to Jboss they have made it easy to integrate with VLEs? What I did not know about was integration of Active Directory. This I did not read about from their sites.

 

I don't care why they are using Jboss and SQL. In fact, I am rather irked by the fact that they chose to only support MS SQL when they could have easily tacked on support for Postgresql and made the whole thing available on Linux/UNIX instead of just Windows whatever Server. Then there is what appears to be their inability to make proper use of the tools that they have chosen. Just because they are using Java and SQL does not automatically mean it will scale.

 

I don't know about you but as far as I am concerned, using a lot of memory is a sign of incompetency and that especially shows when Oliver chews 1GB of RAM and yet cannot perform when it is just, ironic that you should mention 3 users, 3 users and not all are on at the same time too. The kids are not logging on and using OPAC yet and already the thing is a misfit. To add insult to injury, it is running on a box with 16GB of RAM, 12 cores in total and with 12 disks on SAS. You want to tell me that I am simply slagging off a product just because I think it uses too much memory? I have run/managed clusters of servers totalling just 60 boxes that handled over 200 million smtp transactions and 2-5 million deliveries daily. Most of these where just dual PIII boxes while a few where Xeons with a bit more RAM and 9 of these boxes were mysql servers for the clusters. I believe I know what I am talking about when I have had to make crap hardware perform because I ain't getting anything new/better.

 

You are highly mistaken about the part: Oliver used a lot less RAM but then worked poorly. It is: Oliver uses truckloads of RAM and still works poorly.

[feizhou]

As we don't have usernames in SIMS we had to export (from SIMS) admission numbers, names and UPN's - we then exported from AD usernames and UPN's (we hold UPN in AD). we could then form a spreadsheet and merge the two together so that the names, UPN's and usernames matched. (the admission number in Oliver is called STCode). The csv can be created from any field - softlink will be able to support with this.

 

I agree, the software is an absolute pile of sh*te.

 

Thanks, just got my hands on the manual earlier and found out about where CSV templates are defined. The Wisdomlib guy (Softlink's representative in HK) had created a Maze template that had a lot of fields...I also like how they don't tell you to not include field headers...sigh.

 

Oh well, not the only thing 'forced' upon me by HQ. They also torpedoed my asterisk project to replace the school's keyline telephone system but I suppose it should be a plus that I get to play around with Cisco Unified Communications Manager. Another bonus was that I still got to do some asterisk...integrated it with UCM so that I can add sip lines to the system and chop off more ripoff landlines from the local telco monopolist.

[itboss]

well just put Oliver in - recomended server RAM was about 6GB/8GB from memory ?

 

I put in 12 GB to be sure - it works absolutely fine

 

- and everything on that server works at lightning speed now :-)

[matt40k]

Why do you think I would know anything about integration with Moodle IF they did not specifically claim that by moving to Jboss they have made it easy to integrate with VLEs? What I did not know about was integration of Active Directory. This I did not read about from their sites.

 

I don't care why they are using Jboss and SQL. In fact, I am rather irked by the fact that they chose to only support MS SQL when they could have easily tacked on support for Postgresql and made the whole thing available on Linux/UNIX instead of just Windows whatever Server. Then there is what appears to be their inability to make proper use of the tools that they have chosen. Just because they are using Java and SQL does not automatically mean it will scale.

 

I don't know about you but as far as I am concerned, using a lot of memory is a sign of incompetency and that especially shows when Oliver chews 1GB of RAM and yet cannot perform when it is just, ironic that you should mention 3 users, 3 users and not all are on at the same time too. The kids are not logging on and using OPAC yet and already the thing is a misfit. To add insult to injury, it is running on a box with 16GB of RAM, 12 cores in total and with 12 disks on SAS. You want to tell me that I am simply slagging off a product just because I think it uses too much memory? I have run/managed clusters of servers totalling just 60 boxes that handled over 200 million smtp transactions and 2-5 million deliveries daily. Most of these where just dual PIII boxes while a few where Xeons with a bit more RAM and 9 of these boxes were mysql servers for the clusters. I believe I know what I am talking about when I have had to make crap hardware perform because I ain't getting anything new/better.

 

You are highly mistaken about the part: Oliver used a lot less RAM but then worked poorly. It is: Oliver uses truckloads of RAM and still works poorly.

 

You don't make any sense. It's like say my core switch has 1gb RAM and it handles 32gb/s of traffic 24/7. Oliver stores thounsands of books, with large number of cross-ref indexes, search terms and all that, which is going to use RAM because discs are slow. SMTP and POP is peanuts. I wouldn't be suprised to find your using the Express version of MSSQL and your using 1GB and 1 core.

[CyberNerd]

Library databases shouldn't need that sort of system requirements, for years universities have run libraries for hundreds of thousands of records on much more modest resources. Oliver uses way more resources than it needs to.

 

This is free and looks like it's used in some large setups:

Welcome to koha.org! — Koha - Open Source ILS - Integrated Library System

[localzuk]

Our school library has about 3000 books if I recall correctly, we have 600 kids and about 50 staff on it too. One librarian, and a couple of machines which the kids use to search it.

 

I have Oliver on a 2 system setup as recommended. One server has 3GB RAM - this holds the database. The other has 4GB RAM and runs the front-end stuff. Neither gets anywhere near the memory available to them. ie. the database server is hitting around 1.8GB RAM in use on an average day. The front-end hits 2GB at most.

 

Considering the speed at which it returns search results etc... I'm quite happy with that memory use. Complaining about memory use for it is a bit pointless in my mind, considering that RAM is cheap and most people seem to base their views of what is 'good' on decade old ideas about memory usage and resource allocation.

[feizhou]

You don't make any sense. It's like say my core switch has 1gb RAM and it handles 32gb/s of traffic 24/7. Oliver stores thounsands of books, with large number of cross-ref indexes, search terms and all that, which is going to use RAM because discs are slow. SMTP and POP is peanuts. I wouldn't be suprised to find your using the Express version of MSSQL and your using 1GB and 1 core.

 

Oliver stores squat. MSSQL stores the actual data but I don't see it chewing up 1GB of RAM. Yeah, so I am running the express version of MSSQL. I had 2GB and 2 Cores allocated to it initially. Now I have 4GB and 4 Cores allocated to it and it is behaving better. In other words, Oliver is crap. At one point it was chewing almost 1.5GB of RAM. So you want to tell me a dumb frontend to a sql database needs to use 1GB of RAM just right after startup?

 

PicPaste - oliver_chews_1GB_RAM-XPOV6rll.png

 

PicPaste - oliver_needs_4GB_RAM-aJFxykzL.png

 

Like CyberNerd has pointed out, other library systems are way more efficient that this excuse for a sql db frontend called Oliver

[feizhou]

well just put Oliver in - recomended server RAM was about 6GB/8GB from memory ?

 

I put in 12 GB to be sure - it works absolutely fine

 

- and everything on that server works at lightning speed now :-)

 

If I had a choice, we would be using something that runs on UNIX, uses lesser hardware and still run circles around Oliver with the current setup.

[feizhou]

Our school library has about 3000 books if I recall correctly, we have 600 kids and about 50 staff on it too. One librarian, and a couple of machines which the kids use to search it.

 

I have Oliver on a 2 system setup as recommended. One server has 3GB RAM - this holds the database. The other has 4GB RAM and runs the front-end stuff. Neither gets anywhere near the memory available to them. ie. the database server is hitting around 1.8GB RAM in use on an average day. The front-end hits 2GB at most.

 

Considering the speed at which it returns search results etc... I'm quite happy with that memory use. Complaining about memory use for it is a bit pointless in my mind, considering that RAM is cheap and most people seem to base their views of what is 'good' on decade old ideas about memory usage and resource allocation.

 

/me rotfl. Of course they recommend that kind of setup because their design sucks. Oliver would not be able to run with anything less. No RAM, no work. Amazing that it seems to be the only one that demands that kind of resources just to run. 600 kids and 50 staff? We have just a wee bit more at 720 kids and 60 staff. Maybe you can make sense of this: "15232 Bibliographic Resources, 27844 Copies, 8189 Authors, 2423 Subjects, 1120 Publishers, 1713 Series" as to what is stored besides user data. Connecting to the MSSQL database to fix records via python on Linux and running big updates was pretty much instantaneous so any slowness has nothing to do with MSSQL or the hardware. It's that big slow frontend called Oliver that needs uber amounts of RAM. If you are willing to allocate a big box just to run the frontend, fine. I, for one, am not going to allocate anything more than the 4GB of RAM and 4 Cores that I have already done. I ain't asking the school to spend 8,000USD just to run one stupid instance of Oliver.

[feizhou]

Thats pretty much what they said to me, and it was those rumors that lead us to upgrade to the latest version - only to find that it wasn't there and the AD "integration" requires CSV imports and is actually more of a pita than I could have possibly imagined.

 

Hi CyberNerd,

 

I was poking around Oliver and found a whole bunch of parameters that seem to imply AD integration and even single sign on. Running version 4.2 here. Is that what you have too?